5336
J . Org. Chem. 1981,46, 5336-5340
crystallized mp 129 O C ; IR (CHC13)2970,1140cm-'; 'H NMR (CDC13)6 4.10 (2 H,m), 3.85 (1 H, s), 2.6-0.6 (11 H, complex); mass spectrum,m / e (relative intensity) 188/186 (11/36),79 (100). 4-anti-Bromo-l-isopropyl-2-oxaadama~tane (10). A 3 W m g sample of la (1.77 mmol) was treated with thionyl bromide as described for the synthesis of 8. The yield was 118 mg (26%) of 10 as a yellowish oil: IR (CHC13) 2950,1050 cm-'; 'H NMR (CDC13)6 4.55 (1 H, m), 4.04 (1 H, m), 2.6-1.1 (11H, complex), 0.88 (6 H, d); mass spectrum, m / e (relative intensity) 260/258 (15/16,M+), 217/215 (7/7),179 (171,174(51),172 (541, 123 (70), 91 (901,43 (100). 2-anti-Iodo-4,4-dimethyladamantane(23). A 185-mgsample of la (0.93mmol) was refluxed overnight in 57% hydriodic acid.
After cooling the reaction mixture was diluted with water and extracted with methylene chloride. After the usual workup, 149 mg (52%) of 23 as a colorless liquid was obtained which turned dark after some hours at room temperature: IR (CHC13)2950, 1460 cm-'; 'H NMR (CDC13) 6 5.37 (1 H, m), 2.5-1.2 (12 H, 44.6 (e2), complex), 1.10 (6 H, s); 13C NMR (CDC13)6 47.0 (e3), 39.0 (e4),37.4 (C'), 36.7 (e5),34.9 (e8), 34.7 (e9),34.1 (es),28.8 (elo),28.5 (e7),27.5 (BCH,);mass spectrum, m l e (relative intensity) 163 (100)(M+ - I).
2-anti-Bromo-4,4-dimethyladamantane. A 158-mgsample of la (0.81 mmol) was refluxed overnight in 48% hydrobromic acid. After cooling, the reaction mixture was diluted with water and extracted with methylene chloride. The usual workup afforded 33 mg (16%) of 2-anti-bromo-4,4-dimethyladamantane as a colorless liquid: IR (CHC13) 2950, 1460 cm-'; 'H NMR (CDC1,) 6 5.12 (1 H, m), 2.5-1.2 (12H, complex), 1.10 (6 H, s); 13C NMR (CDC13)6 61.7 (e2),45.7 (e3), 38.9 (e4),36.4 (C6),36.1
(e'),34.6 (es),33.6 (C6),33.2 (C'),
28.1 (C"), 27.4 (2CH3),27.1 (C'); mass spectrum, m / e (relative intensity) 163 (100)(M+- Br). Adamantane-2,l-dione (17). A 200-mg sample of IC (1.10 mmol) were dissolved in 20 mL water and 25 mL concentrated
sulfuric acid. The mixture was refluxed overnight, and after it cooled, 50 mL water was added. The usual workup with chloroform afforded 50 mg (22%) of 17 as a white solid: mp 282-283 "C; IR (CHC13)2940,2860,1730,1700 cm-'; 'H NMR (CDC13) 6 3.40 (1 H, m), 2.76 (2H, m), 2.5-1.5 (9H, complex); 13CNMR (CDC13)6 207.6 (C2/4),68.3 (e3), 45.0 (@I6), 44.0 (elo), 38.2 30.0 (es),26.0 (C'); mass spectrum, m / e (relative intensity) 164 (100,M+), 136 (141, 79 (78). Anal. Calcd for C1$Il2O2:C, 73.14;H, 7.37. Found C, 73.08; H, 7.54. The 13C NMR data of 4-oxa-5-homoadamantane derivatives will be published in a forthcoming paper.
Acknowledgment. We are indebted to Professor Dr. E. Wenkert for helpful discussions. We are also grateful to Dr. D. Miiller (Bochum) for recording the mass spectra as well as to Mrs. E. Sauerbier for her skillful assistance in some syntheses. This work was supported by the Fonds der Chemischen Industrie. Registry No,la, 66483-57-6;Ib,66483-52-1;IC, 79499-77-7; 6, 78726-35-9; 7, 78726-38-2; 8,78726-36-0; 9a,79499-78-8; 9b,7949979-9;10,78726-37-1; 128,79499-80-2; 17,19214-00-7;19a,79548-72-4; 19b,79499-81-3; 19~, 56820-19-0; 21,18971-91-0; 23,79499-82-4; 24, 56781-86-3; 25,56781-85-2; endo-bicyclo[3.3.1]non-6-ene-3-carboxylic acid, 21932-98-9;2-anti-bromo-4,4-dimethyladamantane, 79499-83-5.
Vinyl Cations. Comparison of Gas-Phase Thermodynamic and Solvolysis Data with ab Initio MO Calculations Herbert Mayr,* Reinhard Schneider, Dieter Wilhelm, and P. v. R. Schleyer Institut fur Organische Chemie der Friedrich-Alexander- Uniuersitiit Erlangen-Niirnberg, 0-8520 Erlangen, Federal Republic of Germany Received July I , 1981
Ab initio MO calculations with the STO-3G basis set level on cyclic and acyclic methyl- and phenyl-substituted vinyl cations have been used, in combination with the experimental heat of formation of the parent vinyl cation, to evaluate AHf' values for a set of vinyl cations. Comparison of these thermodynamic values with solvolysis data shows that only one-quarter of the stabilizing influence of substituents is effective in the solvolysis transition states. The slow solvolysis rates of vinyl compounds are thus not primarily due to the low stability of vinyl cations but to an unusually high kinetic barrier between vinyl derivatives and the corresponding ionic intermediates.
A knowledge of relative carbenium ion stabilities is essential for the interpretation of reactions involving carbocations and is being used to help design syntheses via cationic intermediates.' In principle, such relative stabilities may be obtained from solvolysis data or from thermochemical studies. However, only a few gas-phase heats of formation have been determined. Utilization of reactivity data in solution requires that solvolysis transition states resemble the corresponding carbenium ions energetically. Furthermore, published solvolysis rates have been measured under different conditions with varying leaving groups so that a consistent set of data is not available. Therefore, we have now used molecular orbital calculations to gather such thermochemical information. (1) (a) M a y , H. Angew. Chem.,Int. Ed. Engl. 1981,20,184.(b) Mayr, H.; SchGtz, F. Tetrahedron Lett. 1981, 925. (c) Mayr, H.; Seitz, B.; Halberstadt-Kauach, I. K. J. Org. Chem. 1981,46,1041.
Earlier computational treatments of vinyl cations concentrated on structures and the relative stabilizing abilities of a or p substituents.2 We now extend this work on polysubstituted and cyclic systems in order to obtain heats of formation of vinyl cations with widely varying substitution patterns (1-13). Calculations were carried out at the restricted Hartree-Fock level by using the ab initio SCF-MO GAUSSIAN 70 series of program^.^ Partial geometry optimizations (2)(a) Radom, L.; Hariharan, P. C.; Pople, J. A.; Schleyer, P. v. R. J . Am. Chem. SOC.1973,95,6531.(b) Apeloig, Y.;Schleyer, P. v. R.; Pople, J. A. J. Org. Chem. 1977,42,3004. (c) Apeloig, Y.;Schleyer, P. v. R.; Pople, J. A. J. Am, Chem. SOC.1977,99,5901. (d) Apeloig, Y.;Collins, J. B.; Cremer, D.; Bally, T.; Haselbach, E.; Pople, J. A.; Chandrasekhar, J.; Schleyer, P. v. R. J. Org. Chem. 1980, 45, 3496. (e) Krishnan, R.; Whiteside, R. A.; Pople, J. A.; Schleyer, P. v. R. J. Am. Chem. SOC.1981, 103,5649. (3)Hehre, W.J.; Lathan, W. A.; Ditchfield, R.; Newton, M. D.; Pople, J. A,; QCPE Program No. 236, Indiana University Bloomington, IN.
0022-326318111946-5336$01.25/00 1981 American Chemical Society
J. Org. Chem., Vol. 46, No. 26, 1981
Vinyl Cations
5337
Table I. S T O - 3 G O p t i m i z e d S t r u c t u r e s of Cyclic Vinyl C a t i o n s 9-1 1 in C o m p a r i s o n w i t h 4
+
0 c1clc4
CIC,C,
angle, d e g ; bond l e n g t h , A 2-butenyl (4)
1-cyclohexenyl
1-cyclopentenyl
(91
(10)
155.6 106.8 100.1 1.270 1.4 57 1.577
140.7 93.3 89.2 1.274 1.483 1.587
179.0 125.8
c,c,c, CIC, c,c4
1.284 1.481 1.543
ClC,
involving the heavy-atom framework and the CCH angles were performed with the STO-3G minimal basis set4 by using analytically calculated gradientsq5v6Complete optimizations were carried out for the linear systems with less than five carbon atoms.'
Results and Discussion Geometry. Acyclic vinyl cations prefer linear arrangements a t the positive carbon; for systems in which exact linearity is not allowed by symmetry, deviations of less than loare calculated. The C=C bond lengths are practically unaffected by substitution (1.281-1.286 A). This value for 1 is 1.262 A with the much larger 6-31G* basis. 2e For 7 and 8, the previously calculated minimum energy conformation with a perpendicular phenyl ring2bwas enforced by symmetry constraints. Optimization of the benzene ring of 7 resulted in elongation of the Ci-C, (1.440 A) and the C,-C, bonds (1.407 A) and in shortening of the C,-C, bond (1.372 A); an energy gain of 2.9 kcal/mol relative to the partially optimized structure2bresulted. The benzene ring geometry is closely similar to that calculated for the benzyl cation.8 In agreement with findings for the fl-methylvinylcation 3," the &@-dimethylvinylcation 5 prefers the conformation
\
H -C
t
C=C-H
C=C-H
/
&
HH
*
'
\
+
.C =C
-H
8,
(+ 2.6 kcal/mol)
(4.1 kcal/mol)
( 0.0 kcal/mol)
with the cis hydrogens in the plane of the empty p orbital. Consequently, similar conformations were selected for 6 and 8. Eclipsed conformations were chosen for the amethyl groups in 4 and 6. Such conformations were calculated to be slightly more stable than the perpendicular arrangement in cation 2.2a The geometries of the cyclic vinyl cations 9-1 1 reflect the tendency of vinyl cations to adopt linear arrangements. Unusually large C2C1C4angles, almost complete flattening of the rings, and decrease of the adjacent C1C2C3and C1C4C5angles result (Table I). In the cyclohexenyl and cyclopentenyl cations, 9 and 10, the ClC2double bond is slightly reduced in length relative to that in 4 whereas C2C3 is considerably elongated (+0.04 A) in both cases. These trends continue in the l-cyclo(4)Hehre, W. J.; Ditchfield, R.; Stewart, R. F.; Pople, J. A. J. Chem. Phys. 1969,51, 2647. (5)Schlegel, H. B.; Wolfe, S.;Bernardi, F. J. Chem. Phys. 1975,63, 3632. (6)Optimization criterion: 6 f / 6 x < 0.01 or until the total energy was constant to better than 0.05 kcal/mol. (7)For the extent of optimization see the footnotes of Table 11. (8) Wolf, J. F.; Harch, R. G.; Taft, R. W.; Hehre, W. J. J.Am. Chem. SOC.1975,97,2902.
l-cyc1obutenylzd (11)
124.2 71.5 1.250 1.479 1.733
butenyl cation 11, which is indicated to have a partially opened (bridged) structure.2d Energy. Inherent deficiencies of the computational methods are most efficiently eliminated or reduced by employing isodesmic reactions. Thus, reliable energy differences of various carbeniqm ions may be obtained from high level ab initio MO calculations.2" Naturally, this procedure is limited to small systems. However, if the isodesmic reactions considered involve only one class of carbenium ions, even from STO-3G minimal basis set data, reliable energy values can be obtained. Equations 1 and 2 demonstrate the coincidence of calculated and experimental AE values of methyl and phenyl transfer reactions in alkyl cation^.^ (CH3)2CH+
+ C2H6
-
(CHJ3C'
+ CH4
(1)
hE(ST0-3G) = -20.2 kcal/mol"; ilE(expt1) = -20.7,12 -22.013 kcal/mol CH3+
+ Ph-CHs
-
Ph-CH2+
+ CH4
(2)
hE(ST0-3G) = -79.0 kcal/mol14; AE(expt1) = -79.9 kcal/mol15 In addition, we have reported that there is an excellent agreement between calculated (STO-3G) and experimental AI3 values for methyl-transfer reactions in the series of methyl substituted allyl cations.l0 Therefore, we believe the stabilization energies of eq 3-11, which are based on the total STO-3G energies of Table 11, will also be reliable. a Substitution HZC=C+H 1
+ C2H6
-
HZC=C+CH3 + CHI 2 hE = -30.3 kcal/mol (3)
(9)Methyl-transfer reactions involving C2H6+are not considered in this context because of the problems arising with the nonclassical structure of the ethyl cation (see reference 24. (10)Mayr, H.; Fbrner, W.; Schleyer, P. v. R. J. Am. Chem. SOC.1979, 101,6032; 1980,102,3663. (11)Total STO-3G energy for (a) isopropyl cation (-116.02765au): Radom, L.;Pople, J. A.; Buss, V.; Schleyer, P. v. R. J. Am. Chem. SOC. 1972,94,311. (b) For ethane (-78.306 18 au) and methane (-39.72686 au): Lathan, W. A.; Hehre, W. J.; Pople, J. A. J. Am. Chem. SOC. 1971, 93, 808. (c) For tert-butyl cation (-154.63918au): Hehre, W.J. In 'Applications of Electronic Structure Theory"; Schaefer, H. F., 111,Ed.; Plenum Press: New York, 1977;Vol. 111. (12)AHfo for isopropyl (192 kcal/mol) and tert-butyl cation (169 kcal/mol): Lowing, F. P.; Semeluk, G. P. Can. J. Chem. 1970,423,955. (13)AHto for isopropyl (187.3kcal/mol) and tert-butyl cation (162.9 kcal/mol): Houle, F. A.; Beauchamp, J. L. J. Am. Chem. SOC.1979,101, 4067. (14)Total STO-3G energy for (a) CH3+ (-38.77948 au) and CHI: ref llb. (b) For toluene (-266.47382au): Hehre, W. J.; Radom, L.; Pople, J. A. J.Am. Chem. SOC.1972,94,1496.(c) For PhCH2+(-265.65231 au): ref llc. (15)AHfofor (a) methyl cation (261kcal/mol): ref 12. (b) For benzyl cation (211kcal/mol): Houle, F. A.; Beauchamp, J. L. J.Am. Chem. SOC. 1978,100,3290.
J. Org. Chem., Vol. 46, No. 26, 1981
5338
H2C=C+H 1
+ Ph-CH3
H&=C+Ph + CH, 7 A,??= -65.5 kcal/mol (4)
-+
-
+ CzH6
(CH3)&=C+H 5 (CH3)2C=C+H
+ PhCH3
5
Mayr et al.
(CH3)&=C+CH3 + CH4 6 AE = -25.5 kcal/mol (5)
-
(CH3)2C=C+Ph + CH4 8 AE = -51.9 kcal/mol (6)
Equations 3 and 4 show that the stabilizing effect of phenyl is approximately twice that of methyl. A similar ratio is found in saturated carbenium ions.16 From eq 5 and 6 it is evident that the stabilizing effect of a sub0 Substitution
1
stressed by the finding that the isodesmic reaction comparing 2 and 11 displays considerable basis set dependence.2d Eqs 16 and 17 show that methylation of the 0-carbon has approximately the same effect in cyclic vinyl cations as in acyclic species (eq 15). H, + ,c=C-CHx CH3 4
C?H,
-
CH, lCd-CH3 CH3
+
CH,
-10.8 kcal/mol (15)
3 AE
=
-15.C kcal/mol
(7) +
4 A € = -12 3 k c a l / m o l ( 8 ) t
HzC=C-H
t ZC2H6
1
-
H C
t 2CH4
\C&-H H3C
/
5 A € = -27 8 kcal/mol ( 9 ) t
HzC=C-CH3
C
ZCzH6
4
H3C\ t /C=C-CH3
t ZCH4
H3C
2
6 A € = - 2 3 1 k c a l / m o l (IO)
t
HzC=C-Ph
7
t ZCzH6
-
H3C\ H3C
/
t
C=C-Ph
t CH4
8 A€
= -17 2 kcal/mol
(11)
stituents is somewhat reduced if the 6-carbon is methyl substituted. Methyl groups at the @-carbonstabilize vinyl cations by 15 (eq 7) and 12 kcal/mol (eq a), though this position is formally uncharged. This effect has been attributed to hyperconjugation.2c As expected, the 6-methylation effect decreases the more the positive charge is stabilized by a substituents (eq 9-11). Cyclic Vinyl Cations.18 Cyclic vinyl cations are less stable than the acyclic analogues. According to eq 12 and 13, 1-cyclohexenyland 1-cyclopentenyl cations are destabilized by 17 and 27 kcal/mol relative to their acyclic counterpart, 4. With a C=C+-C angle of 155O, the 1cyclohexenyl cation is not very different from acyclic species. Therefore, eq 12 can be expected to give a reliable A E value for the reasons discussed above. In the isodesmic reactions, 13 and 14, carbenium ions of widely different structures are correlated, however, and we cannot be sure that the calculated numbers are accurate. This fact is (16) Streitwieser, A., Jr. ‘Solvolytic Displacement Reactions”; McGraw-Hill: New York, 1962; pp 42, 43. (17) Since the phenyl ring of 8 is not optimized, the total energy (-303.012 00 au) of partially optimized 72bwas used for this equation. (18) Total STO-3G energies of (a) cyclohexene (-230.26065 au) and cyclopentene (-191.673 62 au): complete optimization except CH bond lengths. (b) For cyclobutene (-153.040 28 au) and cis-butene (-154.241 09 au): Hehre, W. J.; Pople J. A. J . Am. Chem. SOC.1975,97, 6941.
Heats of Formation. The calculated AE values of eq 3-17 may be used to evaluate heats of formation of vinyl cations 2-13 on the basis of the experimental AH; values of the neutral hydrocarbonslg and of the unsubstituted vinyl cation 1.20 Comparison of the resulting calculated AH; values (Table 11) with the few available experimental data (in kilocalories/mole; see below) show satisfactory agreement. Our assumptions are verified. A H ; (calcd)
A H ; (exptl)
H,C=C+CH, 233 23020,2312‘ CH,CH=C+CH, 218 21520,21822 (CH,),C=CCH, 205 20222 Comparison with Solvolysis Data. On the basis of the Hammond-Leffler postulate,23one can expect a linear correlation between activation free enthalpies of solvolysis reactions (AG’) and the dissociation free enthalpies (AG). If no specific solvation effects exist, the AG* (solvolysis) values should also be related to the gas-phase stabilities of the corresponding vinyl cations (Table 111). Figure 1 shows the linear correlation between the solvolysis free enthalpies of vinyl triflates and the heat of (19) Cox, J. D.; Pilcher, G. “Thermochemistry of Organic and Organometallic Compounds”; Academic Press: New York, 1970. (20) Aue, D. H.; Bowers, M. T. In ‘Gas Phase Ion Chemistry”; Bowers, M. T., Ed.; Academic Press: New York, 1979; Vol. 2, Chapter 9. (21) Aue, D. H.; Davidson, W. R.; Bowers, M. T. J. Am. Chem. SOC. 1976,98,6700. (22) Stang, P. J.; Rappoport, 2.; Hanack, M.; Subramanian, L. R. ‘Vinvl Cations”: Academic Press: New York. 1979: v 7. (5) Leffler, J. E.; Grunwald, E. “Rates and EqGilibria of Organic Reactions”; Wiley: New York, 1963; pp 156, 163. (24) Summenille, R. H.; Senkler, C. A.; Schleyer, P. v. R.; Dueber, T. E.; Stang, P. J. J. Am. Chem. SOC. 1974,96,1100. (25) Pfeifer, W. D.; Bahn, C. A.; Schleyer, P. v. R.; Bocher, S.; Harding, 1971,93, C. E.; Hummel, K.; Hanack, M.; Stang, P. J. J. Am. Chem. SOC. 1513. (26) Hanack, M.; Schleyer, P. v. R.; Martinez, A. G. An. Quim. 1974, 70,941. (27) Benson, S. W. ‘Thermochemical Kinetics”, 2nd Ed.; Wiley-Interscience: New York, 1976. (28) Stang, P. J.; Hargrove, R. J.; Dueber, T. E. J . Chem. Soc., Perkin Trans. 2 1977,1486. (29) Hargrove, R. J.; Dueber, T. E.; Stang, P. J. J . Chem. Soc., Chem. Commun. 1970,1614. (30) (a) Grob, C. A,; Cseh, G. Helu. Chim. Acta 1964,47, 194. (b) Derocque, J.-L; Sundermann, F.-B.; Youssif, N.; Hanack, M. Justus Liebigs Ann. Chem. 1973,419. (31) Hanack, M.; Bentz, H.; Markl, R.; Subramanian, L. R. Justus Liebigs Ann. Chem. 1978,1894.
J. Org. Chem., Vol. 46, No. 26, 1981 5339
Vinyl Cations
Table 11. Total Energies (au, STO-3G) of Vinyl Cations and Calculated Heats of Formation (kcal/mol) R2 'C=C--R
+
I
3/
R
cation
R' H CH3 H CH 3
1 2 3
R'
R3 H H H H CH, CH3 H CH3
I-I
total energy -76.165 -114.792 -114.768 -153.391 -153.368 -191.988 -303.016 -380.197 -229.384 -190.780 -152.155 -267.981 -229.380
40 96 54 86 36 34 67 98 33 94 81 44 56
degree of optimization" I
ref llb 2a
AH; (calcd) 266' 233 249 218 234 205 230 209 23 6 (255)d ( 280) 222 (240)d
I H I 2a CH3 I 4 CH3 I 5 H CH3 I 6 CH 3 CH3 I1 I Ph H 111 8 Ph CH3 9 1-cyclohexenyl IV 10 1 -cyclopentenyl IV 11 1-cyclobutenyl I 2d V 12 2-methyl-1-cyclohexenyl V 13 2-methyl-1-cyclopentenyl a I = complete optimization; I1 = heavy-atom framework fully optimized; I11 = (CH3),C=C+fragment as calculated for 6, with phenyl as reported in ref 2d; IV = complete optimization except for C-H bond lengths; V = geometries of 9 and 10, respective, methyl groups with standard geometry. Complete optimization in C,, symmetry except for aromatic C-H results in an energy gain of 2.9 kcal/mol relative to the structure with phenyl in standard geometry.2b Experimental reference value. d Uncertain; see text. Table 111. Solvolysis Rates of Vinyl Triflates in 50% Ethanol-H,O at 75 "C solvolyses A Hf" k , , s-' AG*, kcal/mol ref A H f " (ion) (precursor)"
R-OTf O ", hZC='
AH^
3.3 x 10-4
26.0
24
23 3
-34.3
267
1.75 x 10-4
26.4
24
218
-41.2
2 59
2.94 x 10-4
26.0
24
20 5
-49.7
255
21.3
b
230
-3.6
234
20.0
C
20 9
-18.9
228
32.6
25
23 6
-39.8
276
25, d
(255)
-30.4
(285)
27.0
e
(280)
-1.5
(282)
30.8
25
-\CH3 H\"
_ _/ C T f
H3C'--"\C
b3
O , T' (CH&C =C 'CH3
/OTf
2.74
@=c\
X
lo-'
Ph OTf
ICH312C I
C/ \P"
1.68
>35.5
6
O
T
f
>31.3
26, d
222
-48.2
270
(240)
-38.8
(279)
" Calculated from group increments;" since the absolute value is not of interest, OTf has been replaced by OCH, for simFrom k , (80%ethanol)" by using the " m y equation" (ref 38), m = 0.568 from ref 29. ' From rate ratio plicity. (CH,),C=C(Br)Ph/H,C=C(Br)Ph at 120 0C.3" Since solvolysis has been shown to proceed via S-0 cleavage3' this value represents a lower limit. e From relative rates of 1-cyclobutenyl and 1-cyclohexenyl nonaflates (see ref 22, p 242). f A H= AHf"(ion)- AHf"(precursor)(ref 38). formation differences between the cations and the precursors. However, two points deviate, those for 2-propenyl and for 1-cyclobutenyl triflates. Since the deviation of cyclobutenyl triflate can be attributed to the inadequacy of our computational approach (see discussion above), 11 has been omitted from the graph. On the other hand, the displacement of 2-propenyl triflate must have a physical meaning; the calculated heat of formation of vinyl cation 2 is in agreement with the experimental gas-phase value. The finding that 2-propenyl triflate solvolyses faster than expected from the gas-phase stability of cation 2 supports a previous suggestion that systems with 0-hydrogens trans to the leaving group experience specific solvation effects
in the solvolysis transition states.% To eliminate this effect in 4, the solvolysis data of (E)-2-buten-2-y1triflate and not of its 2 isomer has been included to the plot. From the correlation line one can expect activation free enthalpies for the formation of cyclopentenyl cations which are not very different from those observed for 0-S cleavage in 1-cyclopentenyl t r i f l a t e ~ . ~Probably, ~ , ~ ~ 0-S cleavage is only slightly favored over the formation of vinyl cations. This assumption is supported by a recent report that 1cyclopentenyl cations are formed readily via R cyclization.lc (32) This conclusion is based on the assumption that STO-3G calculations are still appropriate for 1-cyclopentenyl cations.
J. Org. Chem. 1981,46, 5340-5343
5340
shows that tert-butyl and a-phenylvinyl cations have similar hydride affinities (Le., similar thermodynamic stabilities) though tert-butyl bromideu solvolyses ten powers of ten faster than a - b r o m ~ s t y r e n e .Instead, ~~~ the transition states for vinyl solvolyses reflect the energies of the fully developed ions to a much smaller extent than do the transition states in alkyl solvolyses. Thus, vinyl cations are formed much more slowly than trivalent carbenium ions of equal thermodynamic stability. This point is illustrated nicely by the fact that progargyl bromide 14
Figure 1. Correlation between activation free enthalpies of solvolysis reactions of vinyl triflates and gas-phase stabilities of vinyl cations.
If the questionable data for 2-propenyl triflate and 1cyclobutenyl triflate are omitted, regression analysis yields a linear correlation (r2= 0.980) with a slope of 0.26. This shows that only a small fraction of carbenium ion character is developed in the solvolysis transition states. Corresponding plots for solvolyses of allylic chlorides in formic acid and of tertiary alkyl chlorides in ethanol gave slopes of 0.51'O and 0.66,33 respectively, indicating a higher cationic character in these transition states. Conclusions The slow solvolysis rates of vinylic compounds have been considered to indicate low stability of vinyl cations. This interpretation may be erroneous, however, since eq 18
+
-
Ph-C+=CHZ (CHJBCH 230 kcal/mol -32.4 kcal/mol Ph-CH=CH2 + (CH3)3C+ AH = +1 kcal/moP5 35.3 kcal/mol 163 kcal/mol (18) (33) Schleyer, P. v. R.; M a y , H., unpublished.
HC-CCH,Br 14
+
+
HC-C%H,
+-
BrCH=C=CH, 15
solvolyses 4000 times faster than allenyl bromide 15 though both give the same carbenium ion.% Ground-state effects cannot be responsible since 14 and 15 have similar thermodynamic stabilitie~.~'We suspect that the transition states for vinyl solvolyses are less effectively solvated even when solvent "ionizing power" rather than solvent nucleophilicity is involved. Alternatively, there may be a larger barrier in going from a sp2ground state to an sp ion pair transition state than that for a corresponding sp3sp2 change. Acknowledgment. We thank the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie for financial support and the staff of the Regionales Rechenzentrum Erlangen for their cooperation. Registry No. 1, 14604-48-9; 2, 50457-57-3; 3, 50457-58-4; 4, 79449-26-6; 5, 79449-27-7; 6,79449-28-8; 7,62698-27-5; 8,56126-11-5; 9, 60690-49-5; 10, 60690-50-8; 12, 79449-29-9; 13, 79449-30-2. (34) Fainbere. A. H.: Winstein. S. J.Am. Chem. SOC.1957, 79, 1602. (35) For AH;'(t-Bu+) = 163 kcd/mol;13if AHfo (t-Bu+)is taken to be 169 kcal/mol,'l AH (eq 18) = 7 kcal/mol. (36) Lee, C. V.; Hargrove, R. J.; Dueber, T. E.; Stang, P. J. Tetrahedron Lett. 1971, 2519.(37) Eauilibrium mixture contains 35% 14 and 65% 15: Jacobs, T. L.;'Brill, W. F. J.Am. Chem. SOC.1953, 75, 1314. (38) Grunwald, E.; Winstein, S. J. Am. Chem. SOC.1948, 70, 846.
Active Metals from Potassium-Graphite.' Highly Dispersed Nickel on Graphite as a New Catalyst for the Stereospecific Semihydrogenation of Alkynes Diego Savoia, Emilio Tagliavini, Claudio Trombini, and Achille Umani-Ronchi* Zstituto Chimico "G. Ciamician", Universitd di Bologna, 40126 Bologna, Italy
Received J u l y 3, 1981
The reduction of NiBr2.2DMEby means of potassium-graphite affords highly dispersed nickel on the graphite surface (Ni-Grl), Freshly prepared Ni-Grl is used "in situ" as a catalyst for semihydrogenation of alkynes to alkenes in the presence of ethylenediamine as catalyst modifier. Unconjugated and conjugated (a-alkenes with a stereospecificity of 96-99% and 94%, respectively, are obtained.
Alkynes are unreplaceable tools in organic chemistry for the facile construction of carbon-carbon bonds through acetylides and for the reduction to ( E ) -or (2)-alkenes by a variety of stereospecific methods. (1)Previous works on the use of potassium-graphite in organic synthesis: Savoia, D.; Trombini, c.;Umani-Ronchi, A. J.Chem. Soc., Perkin Trans. 1 1977,123; Tetrahedron Lett. 1977,653; J . Org. Chem. 1978,43, 2907. Contento, M.; Savoia, D.; Trombini, C.; Umani-Ronchi, A. Synthesis 1979, 30.
0022-3263/81/1946-5340$01.25/0
Heterogeneous catalytic hydrogenation is a topic in organic synthesis,2 especially in the field of natural products, since it allows alkynes to be used as masked (Z)-alk e n e ~ . For ~ this purpose the most widely used catalysts consist in palladium dispersed on suitable supports, the Lindlar catalyst affording the best stereospecificity, since (2) Rylander, P. N. 'Catalytic Hydrogenation in Organic Synthesis"; Academic Press: New York, 1979; Chapter 2. (3) Henrick, C. A. Tetrahedron 1977, 33, 1845.
0 1981 American Chemical Society
